Learning to Transfer with von Neumann Conditional Divergence

TL;DR

This paper introduces von Neumann conditional divergence to enhance domain adaptation by better capturing the relationship between features and labels, leading to improved transferability and reduced forgetting.

Contribution

It proposes a novel divergence measure that quantifies feature-label dependence and integrates it into transfer learning frameworks for multiple source tasks.

Findings

Achieves smaller generalization error on new tasks

Reduces catastrophic forgetting in sequential learning

Outperforms state-of-the-art methods

Abstract

The similarity of feature representations plays a pivotal role in the success of problems related to domain adaptation. Feature similarity includes both the invariance of marginal distributions and the closeness of conditional distributions given the desired response $y$ (e.g., class labels). Unfortunately, traditional methods always learn such features without fully taking into consideration the information in $y$, which in turn may lead to a mismatch of the conditional distributions or the mix-up of discriminative structures underlying data distributions. In this work, we introduce the recently proposed von Neumann conditional divergence to improve the transferability across multiple domains. We show that this new divergence is differentiable and eligible to easily quantify the functional dependence between features and $y$. Given multiple source tasks, we integrate this divergence to capture discriminative information in $y$ and design novel learning objectives assuming those source tasks are observed either simultaneously or sequentially. In both scenarios, we obtain favorable performance against state-of-the-art methods in terms of smaller generalization error on new tasks and less catastrophic forgetting on source tasks (in the sequential setup).